By) GEORGE BIDDER. 
ontogenetic history, particularly as to whether in Sycon also 
the lining of the efferent system arises from the granular cells 
(with large nuclei) of the larva. Measuring thirty nuclei of each 
tissue, near the osculum of S. raphanus, gave the following 
average diameters : 
Nuclei of collar-cells : : . 215 p. 
», Of cloacal epithelium . é Be 
»» Of dermal epithelium . : 2 Ge. 
The largest cloacal nucleus is 4°7 , and two thirds were over 
2°5 uw; the largest nucleus of a collar-cell is 2°64, and there is 
no other over 2°54. Between these two classes, therefore, the 
difference is very marked ; but on the other hand, three fourths 
of the collar-cell nuclei and dermal nuclei are mutually indis- 
tinguishable as regards size. 
In a borax-carmine preparation of Leucandra aspera all 
cells but the gonocytes showed a nuclear reticulum, with the 
possible exception of two parenchym cells. Both in Sycon 
and Leucandra the gonocytes show the well-known large 
vesicular nucleus with nucleolus. 
Interstitial Substance. 
The interstitial jelly between the collar-cells, the existence of 
which I have never suspected from living preparations, proves 
in these permanent sections under the apochromatic lens to 
have considerable importance. It appears not only in the 
best sections of S. compressum, but also in sections made at 
Naples from 8S. raphanus. In permanent preparations of the 
normal condition it often reaches to the level of the base of 
the collar, as drawn by Dendy for Leucosolenia (14, pl. 8, fig. 3), 
sinking in a tension-curve between the two.cells (cf. cut, a). In 
sections where Sollas’s membrane has been produced, the mem- 
brane is seen uniting the tops of the collars, separated from 
the surface of the jelly by a vacant space, not being, as Lenden- 
feld suggested (10), a misinterpretation of this surface. I 
satisfied myself that in fig. 17 the line is not the outline of a 
jelly, but actual irregular fusion of collars, the effect being 
